Automatic zero point correction methods and apparatus are known. In such methods and apparatus, a signal corresponding to a reference weight is stored in a first storage. Subsequent weight-proportional signals are compared to the stored signal, the difference between the new and the stored signal being compared to a predetermined threshold value. The digital display is changed to correspond to the new value only when the difference is at least as big as the threshold value. Differences smaller than the threshold value are utilized for zero correction purposes.
Such a method is described in U.S. Pat. No. 3,665,169. In this method, differences smaller than the threshold value (first differences) cause the new weighing signal to be stored in the first storage, instead of the reference value. Thus, in the case of differences less than the threshold value, the full difference is immediately effective in changing the zero point. This method and apparatus has the disadvantage that transient oscillations in the weighing signal cause a full and generally unnessary correction.